--- a/MpdlXmlTextServer.py	Fri Feb 10 14:46:39 2012 +0000
+++ b/MpdlXmlTextServer.py	Tue Feb 28 19:10:08 2012 +0100
@@ -1,15 +1,26 @@
-
 from OFS.SimpleItem import SimpleItem
 from Products.PageTemplates.PageTemplateFile import PageTemplateFile 
-from Ft.Xml import EMPTY_NAMESPACE, Parse
-from Ft.Xml.Domlette import NonvalidatingReader
 
-import md5
-import sys
+import xml.etree.ElementTree as ET
+
+import re
 import logging
 import urllib
-import documentViewer
-from documentViewer import getTextFromNode, serializeNode
+import urlparse
+import base64
+
+from SrvTxtUtils import getInt, getText, getHttpData
+
+def serialize(node):
+    """returns a string containing an XML snippet of node"""
+    s = ET.tostring(node, 'UTF-8')
+    # snip off XML declaration
+    if s.startswith('<?xml'):
+        i = s.find('?>')
+        return s[i+3:]
+
+    return s
+
 
 class MpdlXmlTextServer(SimpleItem):
     """TextServer implementation for MPDL-XML eXist server"""
@@ -21,9 +32,7 @@
     
     manage_changeMpdlXmlTextServerForm = PageTemplateFile("zpt/manage_changeMpdlXmlTextServer", globals())
         
-    def __init__(self,id,title="",serverUrl="http://mpdl-test.mpiwg-berlin.mpg.de/mpdl/interface/", serverName=None, timeout=40):
-    #def __init__(self,id,title="",serverUrl="http://mpdl-text.mpiwg-berlin.mpg.de:30030/mpdl/interface/", serverName=None, timeout=40):    
-        
+    def __init__(self,id,title="",serverUrl="http://mpdl-text.mpiwg-berlin.mpg.de/mpdl/interface/", serverName=None, timeout=40):
         """constructor"""
         self.id=id
         self.title=title
@@ -35,428 +44,348 @@
         
     def getHttpData(self, url, data=None):
         """returns result from url+data HTTP request"""
-        return documentViewer.getHttpData(url,data,timeout=self.timeout)
+        return getHttpData(url,data,timeout=self.timeout)
     
     def getServerData(self, method, data=None):
         """returns result from text server for method+data"""
         url = self.serverUrl+method
-        return documentViewer.getHttpData(url,data,timeout=self.timeout)
+        return getHttpData(url,data,timeout=self.timeout)
+
 
-    def getSearch(self, pageinfo=None,  docinfo=None):
-        """get search list"""
-        docpath = docinfo['textURLPath'] 
-        url = docinfo['url']
-        pagesize = pageinfo['queryPageSize']
-        pn = pageinfo.get('searchPN',1)
-        #sn = pageinfo['sn']
-        s = pageinfo['s']
-        highlightElementPos =pageinfo ['highlightElementPos']
-        highlightElement = pageinfo ['highlightElement']
-        
-        highlightQuery = pageinfo['highlightQuery']
-        query =pageinfo['query']
-        queryType =pageinfo['queryType']
-        viewMode=  pageinfo['viewMode']
-        tocMode = pageinfo['tocMode']
-        characterNormalization = pageinfo['characterNormalization']
-        #optionToggle = pageinfo['optionToggle']
-        tocPN = pageinfo['tocPN']
-        selfurl = self.absolute_url()
-        data = self.getServerData("doc-query.xql","document=%s&mode=%s&queryType=%s&query=%s&queryResultPageSize=%s&queryResultPN=%s&s=%s&viewMode=%s&characterNormalization=%s&highlightElementPos=%s&highlightElement=%s&highlightQuery=%s"%(docpath, 'text', queryType, urllib.quote(query), pagesize, pn, s, viewMode,characterNormalization, highlightElementPos, highlightElement, urllib.quote(highlightQuery)))
-        #data = self.getServerData("doc-query.xql","document=%s&mode=%s&queryType=%s&query=%s&queryResultPageSize=%s&queryResultPN=%s&sn=%s&viewMode=%s&characterNormalization=%s&highlightQuery=%s"%(docpath, 'text', queryType, urllib.quote(query), pagesize, pn, sn, viewMode,characterNormalization, urllib.quote(highlightQuery)))
-        pagexml = data.replace('?document=%s'%str(docpath),'?url=%s'%url)
-        pagedom = Parse(pagexml)
-        
-        """
-        pagedivs = pagedom.xpath("//div[@class='queryResultHits']") 
-        if (pagedivs == pagedom.xpath("//div[@class='queryResultHits']")):
-            if len(pagedivs)>0:
-                docinfo['queryResultHits'] = int(getTextFromNode(pagedivs[0]))
-                s = getTextFromNode(pagedivs[0])
-                s1 = int(s)/10+1
-                try:
-                    docinfo['queryResultHits'] = int(s1)
-                    logging.debug("SEARCH ENTRIES: %s"%(s1))
-                except:
-                    docinfo['queryResultHits'] = 0
-        """
-        if (queryType=="fulltext")or(queryType=="xpath")or(queryType=="xquery")or(queryType=="fulltextMorphLemma"):   
-            pagedivs = pagedom.xpath("//div[@class='queryResultPage']")
-            if len(pagedivs)>0:
-                pagenode=pagedivs[0]
-                links=pagenode.xpath("//a")
-                for l in links:
-                    hrefNode = l.getAttributeNodeNS(None, u"href")
-                    if hrefNode:
-                        href = hrefNode.nodeValue
-                        if href.startswith('page-fragment.xql'):
-                            selfurl = self.absolute_url()            
-                            pagexml=href.replace('mode=text','mode=texttool&viewMode=%s&queryType=%s&query=%s&queryResultPageSize=%s&queryResultPN=%s&tocMode=%s&searchPN=%s&tocPN=%s&characterNormalization=%s'%(viewMode,queryType,urllib.quote(query),pagesize,pn,tocMode,pn,tocPN, characterNormalization))
-                            hrefNode.nodeValue = pagexml.replace('page-fragment.xql','%s'%selfurl)                                           
-                #logging.debug("PUREXML :%s"%(serializeNode(pagenode)))
-                return serializeNode(pagenode)        
-        if (queryType=="fulltextMorph"):
-            pagedivs = pagedom.xpath("//div[@class='queryResult']")
-            if len(pagedivs)>0:
-                pagenode=pagedivs[0]
-                links=pagenode.xpath("//a")
-                for l in links:
-                    hrefNode = l.getAttributeNodeNS(None, u"href")
-                    if hrefNode:
-                        href = hrefNode.nodeValue
-                        if href.startswith('page-fragment.xql'):
-                            selfurl = self.absolute_url()       
-                            pagexml=href.replace('mode=text','mode=texttool&viewMode=%s&queryType=%s&query=%s&queryResultPageSize=%s&queryResultPN=%s&tocMode=%s&searchPN=%s&tocPN=%s&characterNormalization=%s'%(viewMode,queryType,urllib.quote(query),pagesize,pn,tocMode,pn,tocPN,characterNormalization))
-                            hrefNode.nodeValue = pagexml.replace('page-fragment.xql','%s'%selfurl)  
-                        if href.startswith('../lt/lemma.xql'):
-                            hrefNode.nodeValue = href.replace('../lt/lemma.xql','%s/template/head_main_query'%(selfurl))        
-                            l.setAttributeNS(None, 'target', '_blank')
-                            l.setAttributeNS(None, 'onClick',"popupWin = window.open(this.href, 'contacts', 'location,width=300,height=400,top=180, left=400, scrollbars=1'); return false;")
-                            l.setAttributeNS(None, 'onDblclick', 'popupWin.focus();')  
-                pagedivs = pagedom.xpath("//div[@class='queryResultMorphExpansion']")                
-                return serializeNode(pagenode)        
-        if (queryType=="ftIndex")or(queryType=="ftIndexMorph"):
-            pagedivs= pagedom.xpath("//div[@class='queryResultPage']")
-            if len(pagedivs)>0:
-                pagenode=pagedivs[0]
-                links=pagenode.xpath("//a")
-                for l in links:
-                    hrefNode = l.getAttributeNodeNS(None, u"href")
-                    if hrefNode:
-                        href = hrefNode.nodeValue
-                        hrefNode.nodeValue=href.replace('mode=text','mode=texttool&viewMode=%s&tocMode=%s&tocPN=%s&pn=%s&characterNormalization=%s'%(viewMode,tocMode,tocPN,pn,characterNormalization))             
-                        if href.startswith('../lt/lex.xql'):
-                            hrefNode.nodeValue = href.replace('../lt/lex.xql','%s/template/head_main_lex'%selfurl)         
-                            l.setAttributeNS(None, 'target', '_blank')
-                            l.setAttributeNS(None, 'onClick',"popupWin = window.open(this.href, 'contacts', 'location,width=500,height=600,top=180, left=400, scrollbars=1'); return false;")
-                            l.setAttributeNS(None, 'onDblclick', 'popupWin.focus();')
-                        if href.startswith('../lt/lemma.xql'):
-                            hrefNode.nodeValue = href.replace('../lt/lemma.xql','%s/template/head_main_lemma'%(selfurl))        
-                            l.setAttributeNS(None, 'target', '_blank')
-                            l.setAttributeNS(None, 'onClick',"popupWin = window.open(this.href, 'contacts', 'location,width=300,height=400,top=180, left=400, scrollbars=1'); return false;")
-                            l.setAttributeNS(None, 'onDblclick', 'popupWin.focus();')
-                return serializeNode(pagenode)      
-        return "no text here"   
-           
-    def getGisPlaces(self, docinfo=None, pageinfo=None):
-        """ Show all Gis Places of whole Page"""
-        xpath='//place'
+    def getPlacesOnPage(self, docinfo=None, pn=None):
+        """Returns list of GIS places of page pn"""
         docpath = docinfo.get('textURLPath',None)
         if not docpath:
             return None
 
-        url = docinfo['url']
-        selfurl = self.absolute_url()
-        pn = pageinfo['current']
-        hrefList=[]
-        myList= ""
-        text=self.getServerData("xpath.xql", "document=%s&xpath=%s&pn=%s"%(docinfo['textURLPath'],xpath,pn))
-        dom = Parse(text)
-        result = dom.xpath("//result/resultPage/place")
+        places=[]
+        text=self.getServerData("xpath.xql", "document=%s&xpath=//place&pn=%s"%(docpath,pn))
+        dom = ET.fromstring(text)
+        result = dom.findall(".//resultPage/place")
         for l in result:
-            hrefNode= l.getAttributeNodeNS(None, u"id")
-            href= hrefNode.nodeValue
-            hrefList.append(href)
-            myList = ",".join(hrefList)
-        #logging.debug("getGisPlaces :%s"%(myList))                             
-        return myList
+            id = l.get("id")
+            name = l.text
+            place = {'id': id, 'name': name}
+            places.append(place)
+
+        return places
     
-    def getAllGisPlaces (self, docinfo=None, pageinfo=None):
-        """Show all Gis Places of whole Book """
-        xpath ='//echo:place'
-        docpath =docinfo['textURLPath']
-        url = docinfo['url']
-        selfurl =self.absolute_url()
-        pn =pageinfo['current']
-        hrefList=[]
-        myList=""
-        text=self.getServerData("xpath.xql", "document=%s&xpath=%s"%(docinfo['textURLPath'],xpath))
-        dom =Parse(text)
-        result = dom.xpath("//result/resultPage/place")
-        
-        for l in result:
-            hrefNode = l.getAttributeNodeNS(None, u"id")
-            href= hrefNode.nodeValue
-            hrefList.append(href)
-            myList = ",".join(hrefList)
-            #logging.debug("getALLGisPlaces :%s"%(myList))
-        return myList
           
-           
-    def getTextPage(self, mode="text_dict", pn=1, docinfo=None, pageinfo=None):
-        """returns single page from fulltext"""
-        docpath = docinfo['textURLPath']
-        path = docinfo['textURLPath']
-        url = docinfo.get('url',None)
-        name = docinfo.get('name',None)
-        pn =pageinfo['current']
-        #sn = pageinfo['sn']
-        s = pageinfo['s']
-        highlightElementPos =pageinfo ['highlightElementPos']
-        highlightElement = pageinfo ['highlightElement']
-        #optionToggle =pageinfo ['optionToggle']
-        highlightQuery = pageinfo['highlightQuery']
-        #mode = pageinfo ['viewMode']
-        tocMode = pageinfo['tocMode']
-        xpointer = pageinfo['xpointer']
-        characterNormalization=pageinfo['characterNormalization']
-        tocPN = pageinfo['tocPN']
-        selfurl = self.absolute_url()   
+    def processPageInfo(self, dom, docinfo, pageinfo):
+        """processes page info divs from dom and stores in docinfo and pageinfo"""
+        # assume first second level div is pageMeta
+        alldivs = dom.find("div")
+        
+        if alldivs is None or alldivs.get('class', '') != 'pageMeta':
+            logging.error("processPageInfo: pageMeta div not found!")
+            return
         
-        if mode == "text_dict":
-            textmode = "textPollux"
-        else:
-            textmode = mode
-        
-        textParam = "document=%s&mode=%s&pn=%s&characterNormalization=%s&xpointer=%s&options=withIdentifier"%(docpath,textmode,pn,characterNormalization, xpointer)
-        if highlightQuery is not None:
-            #textParam +="&highlightQuery=%s&sn=%s"%(urllib.quote(highlightQuery),sn)
-            textParam +="&highlightQuery=%s&s=%s&highlightElement=%s&highlightElementPos=%s"%(urllib.quote(highlightQuery),s, highlightElement, highlightElementPos)           
-        
-        pagexml = self.getServerData("page-fragment.xql",textParam)
-        dom = Parse(pagexml)
-        #dom = NonvalidatingReader.parseStream(pagexml)
-        
-        #original Pages
-        pagedivs = dom.xpath("//div[@class='pageNumberOrig']")       
-        
-        """if pagedivs == dom.xpath("//div[@class='pageNumberOrig']"):
-            if len(pagedivs)>0:
-                docinfo['pageNumberOrig']= getTextFromNode(pagedivs[0])
-                logging.debug("ORIGINAL PAGE: %s"%(docinfo['pageNumberOrig']))
+        for div in alldivs:
+            dc = div.get('class')
+            
+            # pageNumberOrig  
+            if dc == 'pageNumberOrig':
+                pageinfo['pageNumberOrig'] = div.text
+                
+            # pageNumberOrigNorm
+            elif dc == 'pageNumberOrigNorm':
+                pageinfo['pageNumberOrigNorm'] = div.text
+                
+            # pageHeaderTitle
+            elif dc == 'pageHeaderTitle':
+                pageinfo['pageHeaderTitle'] = div.text
+                
+            # numFigureEntries
+            elif dc == 'countFigureEntries':
+                docinfo['numFigureEntries'] = getInt(div.text)
+                
+            # numTocEntries
+            elif dc == 'countTocEntries':
+                # WTF: s1 = int(s)/30+1
+                docinfo['numTocEntries'] = getInt(div.text)
+                
+            # numPlaces
+            elif dc == 'countPlaces':
+                docinfo['numPlaces'] = getInt(div.text)
+                
+            # numTextPages
+            elif dc == 'countPages':
+                np = getInt(div.text)                    
+                if np > 0:
+                    docinfo['numTextPages'] = np
+                    if docinfo.get('numPages', 0) == 0:
+                        # seems to be text-only - update page count
+                        docinfo['numPages'] = np
+                        #pageinfo['end'] = min(pageinfo['end'], np)
+                        pageinfo['numgroups'] = int(np / pageinfo['groupsize'])
+                        if np % pageinfo['groupsize'] > 0:
+                            pageinfo['numgroups'] += 1
         
-        #original Pages Norm
-        pagedivs = dom.xpath("//div[@class='pageNumberOrigNorm']")
-        if pagedivs == dom.xpath("//div[@class='pageNumberOrigNorm']"):
-            if len(pagedivs)>0:
-                docinfo['pageNumberOrigNorm']= getTextFromNode(pagedivs[0])
-                logging.debug("ORIGINAL PAGE NORM: %s"%(docinfo['pageNumberOrigNorm']))
-        """
-        #figureEntries
-        pagedivs = dom.xpath("//div[@class='countFigureEntries']")
-        if pagedivs == dom.xpath("//div[@class='countFigureEntries']"):
-            if len(pagedivs)>0:
-                docinfo['countFigureEntries'] = getTextFromNode(pagedivs[0])
-                s = getTextFromNode(pagedivs[0])
-                if s=='0':
-                    try:
-                        docinfo['countFigureEntries'] = int(s)
-                    except:
-                        docinfo['countFigureEntries'] = 0
-                else:
-                    s1 = int(s)/30+1
-                    try:
-                        docinfo['countFigureEntries'] = int(s1)
-                    except:
-                        docinfo['countFigureEntries'] = 0      
+        #logging.debug("processPageInfo: pageinfo=%s"%repr(pageinfo))
+        return
+         
+           
+    def getTextPage(self, mode="text", pn=1, docinfo=None, pageinfo=None):
+        """returns single page from fulltext"""
         
-        #allPlaces
-        pagedivs = dom.xpath("//div[@class='countPlaces']")
-        if pagedivs == dom.xpath("//div[@class='countPlaces']"):
-            if len(pagedivs)>0:
-                docinfo['countPlaces']= getTextFromNode(pagedivs[0])
-                s = getTextFromNode(pagedivs[0])
-                try:
-                    docinfo['countPlaces'] = int(s)
-                except:
-                    docinfo['countPlaces'] = 0
+        logging.debug("getTextPage mode=%s, pn=%s"%(mode,pn))
+        # check for cached text -- but ideally this shouldn't be called twice
+        if pageinfo.has_key('textPage'):
+            logging.debug("getTextPage: using cached text")
+            return pageinfo['textPage']
+        
+        docpath = docinfo['textURLPath']
+        # just checking
+        if pageinfo['current'] != pn:
+            logging.warning("getTextPage: current!=pn!")
+            
+        # stuff for constructing full urls
+        selfurl = docinfo['viewerUrl']
+        textParams = {'document': docpath,
+                      'pn': pn}
+        if 'characterNormalization' in pageinfo:
+            textParams['characterNormalization'] = pageinfo['characterNormalization']
         
-        #tocEntries
-        pagedivs = dom.xpath("//div[@class='countTocEntries']")
-        if pagedivs == dom.xpath("//div[@class='countTocEntries']"):
-            if len(pagedivs)>0:
-                docinfo['countTocEntries'] = int(getTextFromNode(pagedivs[0]))
-                s = getTextFromNode(pagedivs[0])
-                if s=='0':
-                    try:
-                        docinfo['countTocEntries'] = int(s)
-                    except:
-                        docinfo['countTocEntries'] = 0
-                else:
-                    s1 = int(s)/30+1
-                    try:
-                        docinfo['countTocEntries'] = int(s1)
-                    except:
-                        docinfo['countTocEntries'] = 0
+        if not mode:
+            # default is dict
+            mode = 'text'
+
+        modes = mode.split(',')
+        # check for multiple layers
+        if len(modes) > 1:
+            logging.debug("getTextPage: more than one mode=%s"%mode)
+            
+        # search mode
+        if 'search' in modes:
+            # add highlighting
+            highlightQuery = pageinfo.get('highlightQuery', None)
+            if highlightQuery:
+                textParams['highlightQuery'] = highlightQuery
+                textParams['highlightElement'] = pageinfo.get('highlightElement', '')
+                textParams['highlightElementPos'] = pageinfo.get('highlightElementPos', '')
+                
+            # ignore mode in the following
+            modes.remove('search')
+                            
+        # other modes don't combine
+        if 'dict' in modes:
+            # dict is called textPollux in the backend
+            textmode = 'textPollux'
+        elif len(modes) == 0:
+            # text is default mode
+            textmode = 'text'
+        else:
+            # just take first mode
+            textmode = modes[0]
         
-        #numTextPages
-        pagedivs = dom.xpath("//div[@class='countPages']")
-        if pagedivs == dom.xpath("//div[@class='countPages']"):
-            if len(pagedivs)>0:
-                docinfo['numPages'] = getTextFromNode(pagedivs[0])
-                s = getTextFromNode(pagedivs[0])
-                
-                try:
-                    docinfo['numPages'] = int(s)
-                    #logging.debug("PAGE NUMBER: %s"%(s))
-                
-                    np = docinfo['numPages']
-                    pageinfo['end'] = min(pageinfo['end'], np)
-                    pageinfo['numgroups'] = int(np / pageinfo['groupsize'])
-                    if np % pageinfo['groupsize'] > 0:
-                        pageinfo['numgroups'] += 1                
-                except:
-                    docinfo['numPages'] = 0
-                  
-        else:
-         #no full text -- init to 0
-            docinfo['pageNumberOrig'] = 0
-            docinfo['countFigureEntries'] = 0
-            docinfo['countPlaces'] = 0
-            docinfo['countTocEntries'] = 0
-            docinfo['numPages'] = 0
-            docinfo['pageNumberOrigNorm'] = 0
-        #return docinfo
+        textParams['mode'] = textmode
+        
+        # fetch the page
+        pagexml = self.getServerData("page-fragment.xql",urllib.urlencode(textParams))
+        dom = ET.fromstring(pagexml)
+        # extract additional info
+        self.processPageInfo(dom, docinfo, pageinfo)
+        # page content is in <div class="pageContent">
+        pagediv = None
+        # ElementTree 1.2 in Python 2.6 can't do div[@class='pageContent']
+        # so we look at the second level divs
+        alldivs = dom.findall("div")
+        for div in alldivs:
+            dc = div.get('class')
+            # page content div
+            if dc == 'pageContent':
+                pagediv = div
+                break
         
         # plain text mode
-        if mode == "text":
-            # first div contains text
-            pagedivs = dom.xpath("/div")
-            if len(pagedivs) > 0:      
-                pagenode = pagedivs[0]
-                links = pagenode.xpath("//a")
+        if textmode == "text":
+            # get full url assuming documentViewer is parent
+            selfurl = self.getLink()
+            if pagediv is not None:
+                links = pagediv.findall(".//a")
                 for l in links:
-                    hrefNode = l.getAttributeNodeNS(None, u"href")
-                    if hrefNode:
-                        href= hrefNode.nodeValue
-                        if href.startswith('#note-'):
-                            hrefNode.nodeValue = href.replace('#note-',"?url=%s&viewMode=text&tocMode=%s&tocPN=%s&pn=%s#note-"%(url,tocMode,tocPN,pn))
-                        #if href.startswith():
-                return serializeNode(pagenode)
-        if mode == "xml":
-              # first div contains text
-              pagedivs = dom.xpath("/div")
-              if len(pagedivs) > 0:
-                  pagenode = pagedivs[0]
-                  return serializeNode(pagenode)
-        if mode == "gis":
-              # first div contains text
-              pagedivs = dom.xpath("/div")
-              if len(pagedivs) > 0:
-                  pagenode = pagedivs[0]
-                  links =pagenode.xpath("//a")
-                  for l in links:
-                      hrefNode =l.getAttributeNodeNS(None, u"href")
-                      if hrefNode:
-                          href=hrefNode.nodeValue
-                          if href.startswith('http://mappit.mpiwg-berlin.mpg.de'):
-                              hrefNode.nodeValue =href.replace('db/REST/db/chgis/mpdl','db/RESTdb/db/mpdl/%s'%name)
-                              l.setAttributeNS(None, 'target', '_blank') 
-                  return serializeNode(pagenode)
+                    href = l.get('href')
+                    if href and href.startswith('#note-'):
+                        href = href.replace('#note-',"%s#note-"%selfurl)
+                        l.set('href', href)
+
+                return serialize(pagediv)
+            
+        # text-with-links mode
+        elif textmode == "textPollux":
+            if pagediv is not None:
+                viewerurl = docinfo['viewerUrl']
+                selfurl = self.getLink()
+                # check all a-tags
+                links = pagediv.findall(".//a")
+                for l in links:
+                    href = l.get('href')
                     
-        if mode == "pureXml":
-              # first div contains text
-              pagedivs = dom.xpath("/div")
-              if len(pagedivs) > 0:
-                  pagenode = pagedivs[0]
-                  return serializeNode(pagenode)      
-        # text-with-links mode
-        if mode == "text_dict":
-            # first div contains text
-            #mode = pageinfo ['viewMode']
-            pagedivs = dom.xpath("/div")
-            if len(pagedivs) > 0:
-                pagenode = pagedivs[0]
-                # check all a-tags
-                links = pagenode.xpath("//a")
-                
-                for l in links:
-                    hrefNode = l.getAttributeNodeNS(None, u"href")
-                    
-                    if hrefNode:
+                    if href:
                         # is link with href
-                        href = hrefNode.nodeValue
-                        if href.startswith('http://mpdl-text.mpiwg-berlin.mpg.de/mpdl/interface/lt/wordInfo.xql'):
-                            # is pollux link
-                            selfurl = self.absolute_url()
-                            # change href
-                            hrefNode.nodeValue = href.replace('http://mpdl-text.mpiwg-berlin.mpg.de/mpdl/interface/lt/wordInfo.xql','%s/head_main_voc'%selfurl)
-                            # add target
-                            l.setAttributeNS(None, 'target', '_blank')
-                            #l.setAttributeNS(None, 'onclick',"popupWin = window.open(this.href, 'InfoWindow', 'menubar=no, location,width=500,height=600,top=180, left=700, toolbar=no, scrollbars=1'); return false;")
-                            #l.setAttributeNS(None, "ondblclick", "popupWin.focus();")
-                            #window.open("this.href, 'InfoWindow', 'menubar=no, location,width=500,height=600,top=180, left=700, toolbar=yes, scrollbars=1'"); return false;") 
+                        linkurl = urlparse.urlparse(href)
+                        #logging.debug("getTextPage: linkurl=%s"%repr(linkurl))
+                        if linkurl.path.endswith('GetDictionaryEntries'):
+                            #TODO: replace wordInfo page
+                            # is dictionary link - change href (keeping parameters)
+                            #l.set('href', href.replace('http://mpdl-proto.mpiwg-berlin.mpg.de/mpdl/interface/lt/wordInfo.xql','%s/template/viewer_wordinfo'%viewerurl))
+                            # add target to open new page
+                            l.set('target', '_blank')
                                                           
-                        if href.startswith('http://mpdl-text.mpiwg-berlin.mpg.de/mpdl/lt/lemma.xql'):    
-                            selfurl = self.absolute_url()
-                            hrefNode.nodeValue = href.replace('http://mpdl-text.mpiwg-berlin.mpg.de/mpdl/lt/lemma.xql','%s/head_main_lemma'%selfurl)
-                            l.setAttributeNS(None, 'target', '_blank')
-                            l.setAttributeNS(None, 'onclick',"popupWin = window.open(this.href, 'InfoWindow', 'menubar=no, location,width=300,height=400,top=180, left=700, toolbar=no, scrollbars=1'); return false;")
-                            l.setAttributeNS(None, 'ondblclick', 'popupWin.focus();')   
+                        # TODO: is this needed?
+#                        if href.startswith('http://mpdl-proto.mpiwg-berlin.mpg.de/mpdl/lt/lemma.xql'):
+#                            selfurl = self.absolute_url()
+#                            l.set('href', href.replace('http://mpdl-proto.mpiwg-berlin.mpg.de/mpdl/lt/lemma.xql','%s/head_main_lemma'%selfurl))
+#                            l.set('target', '_blank')
+#                            l.set('onclick',"popupWin = window.open(this.href, 'InfoWindow', 'menubar=no, location,width=500,height=600,top=180, left=700, toolbar=no, scrollbars=1'); return false;")
+#                            l.set('ondblclick', 'popupWin.focus();')   
                     
                         if href.startswith('#note-'):
-                            hrefNode.nodeValue = href.replace('#note-',"?url=%s&viewMode=text_dict&tocMode=%s&tocPN=%s&pn=%s#note-"%(url,tocMode,tocPN,pn))  
+                            # note link
+                            l.set('href', href.replace('#note-',"%s#note-"%selfurl))
                               
-                return serializeNode(pagenode)
-        return "no text here"
-    
-    def getOrigPages(self, docinfo=None, pageinfo=None):
-        docpath = docinfo['textURLPath']
-        pn =pageinfo['current']
-        selfurl = self.absolute_url()   
-        pagexml = self.getServerData("page-fragment.xql","document=%s&pn=%s"%(docpath, pn))
-        dom = Parse(pagexml)
-        pagedivs = dom.xpath("//div[@class='pageNumberOrig']")
-        if pagedivs == dom.xpath("//div[@class='pageNumberOrig']"):
-            if len(pagedivs)>0:
-                docinfo['pageNumberOrig']= getTextFromNode(pagedivs[0])        
-                return docinfo['pageNumberOrig']
-    
-    def getOrigPagesNorm(self, docinfo=None, pageinfo=None):
-        docpath = docinfo['textURLPath']
-        pn =pageinfo['current']
-        selfurl = self.absolute_url()   
-        pagexml = self.getServerData("page-fragment.xql","document=%s&pn=%s"%(docpath, pn))
-        dom = Parse(pagexml)
-        pagedivs = dom.xpath("//div[@class='pageNumberOrigNorm']")
-        if pagedivs == dom.xpath("//div[@class='pageNumberOrigNorm']"):
-            if len(pagedivs)>0:
-                docinfo['pageNumberOrigNorm']= getTextFromNode(pagedivs[0])        
-                return docinfo['pageNumberOrigNorm']
-
-                
-    def getTranslate(self, word=None, language=None, display=None):
-        """translate into another languages"""
-        data = self.getServerData("lt/wordInfo.xql","language="+str(language)+"&word="+urllib.quote(word)+"&display="+urllib.quote(display)+"&output=html")
-        #pagexml=self.template.fulltextclient.eval("/mpdl/interface/lt/lex.xql","document=&language="+str(language)+"&query="+url_quote(str(query)))
-        return data
+                return serialize(pagediv)
+            
+        # xml mode
+        elif textmode == "xml":
+            if pagediv is not None:
+                return serialize(pagediv)
+            
+        # pureXml mode
+        elif textmode == "pureXml":
+            if pagediv is not None:
+                return serialize(pagediv)
+                  
+        # gis mode
+        elif textmode == "gis":
+            if pagediv is not None:
+                # check all a-tags
+                links = pagediv.findall(".//a")
+                # add our URL as backlink
+                selfurl = self.getLink()
+                doc = base64.b64encode(selfurl)
+                for l in links:
+                    href = l.get('href')
+                    if href:
+                        if href.startswith('http://mappit.mpiwg-berlin.mpg.de'):
+                            l.set('href', re.sub(r'doc=[\w+/=]+', 'doc=%s'%doc, href))
+                            l.set('target', '_blank')
+                            
+                return serialize(pagediv)
+                    
+        return None
     
-    def getLemma(self, lemma=None, language=None):
-        """simular words lemma """
-        data = self.getServerData("lt/lemma.xql","language="+str(language)+"&lemma="+urllib.quote(lemma)+"&output=html")
-        return data
-    
-    def getLemmaQuery(self, query=None, language=None):
-        """simular words lemma """
-        data = self.getServerData("lt/lemma.xql","language="+str(language)+"&query="+urllib.quote(query)+"&output=html")
-        return data
-    
-    def getLex(self, query=None, language=None):
-        #simular words lemma
-        data = self.getServerData("lt/lex.xql","document=&language="+str(language)+"&query="+urllib.quote(query))
-        return data
+
+    def getSearchResults(self, mode, query=None, pageinfo=None, docinfo=None):
+        """loads list of search results and stores XML in docinfo"""
+        
+        logging.debug("getSearchResults mode=%s query=%s"%(mode, query))
+        if mode == "none":
+            return docinfo
+              
+        cachedQuery = docinfo.get('cachedQuery', None)
+        if cachedQuery is not None:
+            # cached search result
+            if cachedQuery == '%s_%s'%(mode,query):
+                # same query
+                return docinfo
+            
+            else:
+                # different query
+                del docinfo['resultSize']
+                del docinfo['resultXML']
+        
+        # cache query
+        docinfo['cachedQuery'] = '%s_%s'%(mode,query)
+        
+        # fetch full results
+        docpath = docinfo['textURLPath']
+        params = {'document': docpath,
+                  'mode': 'text',
+                  'queryType': mode,
+                  'query': query,
+                  'queryResultPageSize': 1000,
+                  'queryResultPN': 1,
+                  'characterNormalization': pageinfo.get('characterNormalization', 'reg')}
+        pagexml = self.getServerData("doc-query.xql",urllib.urlencode(params))
+        #pagexml = self.getServerData("doc-query.xql","document=%s&mode=%s&queryType=%s&query=%s&queryResultPageSize=%s&queryResultPN=%s&s=%s&viewMode=%s&characterNormalization=%s&highlightElementPos=%s&highlightElement=%s&highlightQuery=%s"%(docpath, 'text', queryType, urllib.quote(query), pagesize, pn, s, viewMode,characterNormalization, highlightElementPos, highlightElement, urllib.quote(highlightQuery)))
+        dom = ET.fromstring(pagexml)
+        # page content is in <div class="queryResultPage">
+        pagediv = None
+        # ElementTree 1.2 in Python 2.6 can't do div[@class='queryResultPage']
+        alldivs = dom.findall("div")
+        for div in alldivs:
+            dc = div.get('class')
+            # page content div
+            if dc == 'queryResultPage':
+                pagediv = div
+                
+            elif dc == 'queryResultHits':
+                docinfo['resultSize'] = getInt(div.text)
+
+        if pagediv is not None:
+            # store XML in docinfo
+            docinfo['resultXML'] = ET.tostring(pagediv, 'UTF-8')
+
+        return docinfo
     
-    def getQuery (self,  docinfo=None, pageinfo=None, query=None, queryType=None, pn=1):
-         #number of
-         docpath = docinfo['textURLPath'] 
-         pagesize = pageinfo['queryPageSize']
-         pn = pageinfo['searchPN']
-         query =pageinfo['query']
-         queryType =pageinfo['queryType']
-         tocSearch = 0
-         tocDiv = None
-         
-         pagexml = self.getServerData("doc-query.xql","document=%s&mode=%s&queryType=%s&query=%s&queryResultPageSize=%s&queryResultPN=%s"%(docpath, 'text', queryType, urllib.quote(query), pagesize, pn))
-         pagedom = Parse(pagexml)
-         numdivs = pagedom.xpath("//div[@class='queryResultHits']")
-         tocSearch = int(getTextFromNode(numdivs[0]))
-         tc=int((tocSearch/10)+1)
-         return tc
-    
+
+    def getResultsPage(self, mode="text", query=None, pn=None, start=None, size=None, pageinfo=None, docinfo=None):
+        """returns single page from the table of contents"""
+        logging.debug("getResultsPage mode=%s, pn=%s"%(mode,pn))
+        # check for cached result
+        if not 'resultXML' in docinfo:
+            self.getSearchResults(mode=mode, query=query, pageinfo=pageinfo, docinfo=docinfo)
+            
+        resultxml = docinfo.get('resultXML', None)
+        if not resultxml:
+            logging.error("getResultPage: unable to find resultXML")
+            return "Error: no result!"
+        
+        if size is None:
+            size = pageinfo.get('resultPageSize', 10)
+            
+        if start is None:
+            start = (pn - 1) * size
+
+        fullresult = ET.fromstring(resultxml)
+        
+        if fullresult is not None:
+            # paginate
+            first = start-1
+            len = size
+            del fullresult[:first]
+            del fullresult[len:]
+            tocdivs = fullresult
+            
+            # check all a-tags
+            links = tocdivs.findall(".//a")
+            for l in links:
+                href = l.get('href')
+                if href:
+                    # assume all links go to pages
+                    linkUrl = urlparse.urlparse(href)
+                    linkParams = urlparse.parse_qs(linkUrl.query)
+                    # take some parameters
+                    params = {'pn': linkParams['pn'],
+                              'highlightQuery': linkParams.get('highlightQuery',''),
+                              'highlightElement': linkParams.get('highlightElement',''),
+                              'highlightElementPos': linkParams.get('highlightElementPos','')
+                              }
+                    url = self.getLink(params=params)
+                    l.set('href', url)
+                        
+            return serialize(tocdivs)
+        
+        return "ERROR: no results!"
+
+
     def getToc(self, mode="text", docinfo=None):
-        """loads table of contents and stores in docinfo"""
+        """loads table of contents and stores XML in docinfo"""
+        logging.debug("getToc mode=%s"%mode)
         if mode == "none":
-            return docinfo        
+            return docinfo
+              
         if 'tocSize_%s'%mode in docinfo:
             # cached toc
             return docinfo
@@ -472,44 +401,89 @@
         # number of entries in toc
         tocSize = 0
         tocDiv = None
-        
+        # fetch full toc
         pagexml = self.getServerData("doc-query.xql","document=%s&queryType=%s&queryResultPageSize=%s&queryResultPN=%s"%(docpath,queryType, pagesize, pn))
-        
-        # post-processing downloaded xml
-        pagedom = Parse(pagexml)
-        # get number of entries
-        numdivs = pagedom.xpath("//div[@class='queryResultHits']")
-        if len(numdivs) > 0:
-            tocSize = int(getTextFromNode(numdivs[0]))
-        docinfo['tocSize_%s'%mode] = tocSize
+        dom = ET.fromstring(pagexml)
+        # page content is in <div class="queryResultPage">
+        pagediv = None
+        # ElementTree 1.2 in Python 2.6 can't do div[@class='queryResultPage']
+        alldivs = dom.findall("div")
+        for div in alldivs:
+            dc = div.get('class')
+            # page content div
+            if dc == 'queryResultPage':
+                pagediv = div
+                
+            elif dc == 'queryResultHits':
+                docinfo['tocSize_%s'%mode] = getInt(div.text)
+
+        if pagediv is not None:
+            # store XML in docinfo
+            docinfo['tocXML_%s'%mode] = ET.tostring(pagediv, 'UTF-8')
+
         return docinfo
     
-    def getTocPage(self, mode="text", pn=1, pageinfo=None, docinfo=None):
+    def getTocPage(self, mode="text", pn=None, start=None, size=None, pageinfo=None, docinfo=None):
         """returns single page from the table of contents"""
-        # TODO: this should use the cached TOC
+        logging.debug("getTocPage mode=%s, pn=%s"%(mode,pn))
         if mode == "text":
             queryType = "toc"
         else:
             queryType = mode
-        docpath = docinfo['textURLPath']
-        path = docinfo['textURLPath']       
-        pagesize = pageinfo['tocPageSize']
-        pn = pageinfo['tocPN']
-        url = docinfo['url']
-        selfurl = self.absolute_url()  
-        viewMode=  pageinfo['viewMode']
-        characterNormalization = pageinfo ['characterNormalization']
-        #optionToggle =pageinfo ['optionToggle']
-        tocMode = pageinfo['tocMode']
-        tocPN = pageinfo['tocPN']  
+            
+        # check for cached TOC
+        if not docinfo.has_key('tocXML_%s'%mode):
+            self.getToc(mode=mode, docinfo=docinfo)
+            
+        tocxml = docinfo.get('tocXML_%s'%mode, None)
+        if not tocxml:
+            logging.error("getTocPage: unable to find tocXML")
+            return "Error: no table of contents!"
+        
+        if size is None:
+            size = pageinfo.get('tocPageSize', 30)
+            
+        if start is None:
+            start = (pn - 1) * size
+
+        fulltoc = ET.fromstring(tocxml)
         
-        data = self.getServerData("doc-query.xql","document=%s&queryType=%s&queryResultPageSize=%s&queryResultPN=%s&characterNormalization=regPlusNorm"%(docpath,queryType, pagesize, pn))  
-        page = data.replace('page-fragment.xql?document=%s'%str(path),'%s?url=%s&viewMode=%s&tocMode=%s&tocPN=%s'%(selfurl,url, viewMode, tocMode, tocPN))
-        text = page.replace('mode=image','mode=texttool')
-        return text
+        if fulltoc is not None:
+            # paginate
+            first = (start - 1) * 2
+            len = size * 2
+            del fulltoc[:first]
+            del fulltoc[len:]
+            tocdivs = fulltoc
+            
+            # check all a-tags
+            links = tocdivs.findall(".//a")
+            for l in links:
+                href = l.get('href')
+                if href:
+                    # take pn from href
+                    m = re.match(r'page-fragment\.xql.*pn=(\d+)', href)
+                    if m is not None:
+                        # and create new url (assuming parent is documentViewer)
+                        url = self.getLink('pn', m.group(1))
+                        l.set('href', url)
+                    else:
+                        logging.warning("getTocPage: Problem with link=%s"%href)
+                        
+            # fix two-divs-per-row with containing div
+            newtoc = ET.Element('div', {'class':'queryResultPage'})
+            for (d1,d2) in zip(tocdivs[::2],tocdivs[1::2]):
+                e = ET.Element('div',{'class':'tocline'})
+                e.append(d1)
+                e.append(d2)
+                newtoc.append(e)
+                
+            return serialize(newtoc)
+        
+        return "ERROR: no table of contents!"
+    
     
     def manage_changeMpdlXmlTextServer(self,title="",serverUrl="http://mpdl-text.mpiwg-berlin.mpg.de/mpdl/interface/",timeout=40,RESPONSE=None):
-    #def manage_changeMpdlXmlTextServer(self,title="",serverUrl="http://mpdl-text.mpiwg-berlin.mpg.de:30030/mpdl/interface/",timeout=40,RESPONSE=None):
         """change settings"""
         self.title=title
         self.timeout = timeout
@@ -529,4 +503,6 @@
     newObj = MpdlXmlTextServer(id,title,serverUrl,timeout)
     self.Destination()._setObject(id, newObj)
     if RESPONSE is not None:
-        RESPONSE.redirect('manage_main')
\ No newline at end of file
+        RESPONSE.redirect('manage_main')
+        
+        
\ No newline at end of file